Selected research milestones, awards and news from the collaboration’s investigators and partner labs. Broader lab news does not imply Expedition funding; grant support is identified where verified. Dates refer to the original publication or announcement. Reviewed October 2, 2026.
2026
Updated preprint
A parallel probabilistic Ising processor for real-time applications
Cornell, Princeton and USRA collaborators updated their preprint on a densely connected probabilistic Ising machine. First submitted in April, the study connects FPGA hardware evaluation with wireless MIMO detection and acknowledges NSF 1918549.
Modeling loss in broadband quantum optical cavities
Chris Gustin and collaborators develop a cavity-QED treatment for broadband and ultrastrong light–matter interactions. The paper lists Stanford and international affiliations and acknowledges the Expedition grant.
QUBO.jl connects optimization models with Ising solvers
A new journal paper documents the Julia software ecosystem for QUBO reformulation and solver access. Authors span Brazilian institutions, USRA, NASA Ames and Purdue; David E. Bernal Neira acknowledges Expedition support.
The Marandi-led collaboration combines optical memory, linear operations and nonlinear processing. On a waveform task, performance exceeds a linear baseline up to 80 GHz; the paper also credits Fejer and Langrock for a waveguide contribution.
Extending variational quantum optimization to polynomial objectives
A Stanford, Harvard, Caltech and NVIDIA collaboration introduces product-state lifting for higher-order polynomial objectives. The published paper explicitly acknowledges Expedition support for Robin Brown.
Topological soliton frequency combs in lithium niobate
A Caltech-led collaboration reports topological soliton frequency combs in nanophotonic lithium niobate. Published in Nature, the work explicitly acknowledges NSF 1918549.
Studying representation in language-model narratives
Faye-Marie Vassel and collaborators analyze representation and stereotyping in narratives generated by five language models. This grant-supported study broadens the site’s account of the collaboration beyond hardware and optimization.
Programming multimode wave propagation for machine learning
Cornell, NTT Research and Stanford authors demonstrate a reprogrammable two-dimensional waveguide for optical neural-network inference. First published online in December 2025, the paper appears in the January 2026 Nature Physics issue.
Cornell, NTT Research and Stanford researchers report programmable nonlinear optical processing on a chip. The grant-supported paper first appeared online in October 2025 and is collected in Nature’s 2026 volume.
Benchmarking Ising solvers with the cost of tuning included
USRA, NASA Ames, Stanford, Purdue and UCLA authors present an operational benchmarking framework and open-source software. The study measures practical parameter-setting strategies, including their tuning cost, for stochastic optimization solvers.
USRA and Icosa Computing receive NSF STTR Fast-Track award for Combinatorial Reasoning
Icosa Computing and USRA are partners on NSF award 2528318, STTR Fast-Track: Combinatorial Reasoning, with a September 1, 2025 project start date. NSF lists $1,419,331 in funding, with Mert Esencan as PI and Davide Venturelli as co-PI. The project will develop Ising-based optimization methods to select coherent reasoning chains from language-model outputs, aiming to improve AI reliability and efficiency.
NTT and Tohoku advance single-photon Ising-machine research
Yoshihisa Yamamoto’s PHI Lab and Tohoku University reported a journal study of single-photon coherent Ising machines alongside their high-performance-computing collaboration. Numerical simulations found improved solution probabilities on tested clustering problems; physical implementation remained a future step.
Optical neural networks at a few quanta per activation
Cornell, USRA and NTT Research collaborators demonstrate stochastic optical inference with a hidden layer in the single-photon regime. The paper explicitly acknowledges NSF 1918549.
A geometric theory explains how coherent Ising machines find solutions
Atsushi Yamamura, Hideo Mabuchi and Surya Ganguli published a Physical Review X theory of the changing energy landscape inside a coherent Ising machine. Matching theory with numerical experiments, they showed how landscape transitions can guide annealing schedules for random spin-glass problems.
USRA and NASA convene quantum-and-space discussions
Davide Venturelli and USRA colleagues participated in the Quantum World Congress’s Quantum and Space workshop, co-organized by NASA and USRA. Technical talks and breakout sessions connected quantum research with space applications and potential partnerships—a broader community activity involving an Expedition PI.
A joint roadmap for ultrafast quantum nonlinear optics
NTT Research highlighted an Optica perspective bringing together PHI Lab, Stanford and Caltech researchers, including Mabuchi, Fejer and Marandi. The team set out the theoretical and experimental challenges of using stronger optical nonlinearities to bridge classical photonics and quantum information processing.
Topology makes pulsed lasers more resistant to disturbances
Caltech highlighted the Marandi team’s Nature Physics demonstration of topological temporal mode-locking. Coupling the circulating light pulses produces patterns that tolerate certain imperfections and noise, suggesting a route to more robust frequency combs for sensing, communications and computing.
Alireza Marandi receives a DARPA Young Faculty Award
Caltech announced Marandi’s selection for a 2023 DARPA Young Faculty Award for research on few-optical-cycle nonlinear nanophotonic circuits. This separate DARPA award recognizes a research direction closely related to the collaboration’s ultrafast photonics interests.
The Marandi Lab reported a chip-scale mode-locked laser in Science, bringing a building block of ultrafast optics into an integrated platform. Led by Qiushi Guo, the work demonstrated picosecond pulses and tunable repetition rates, opening a route toward more compact photonic systems.
Photonic cellular automata bring a new computing architecture to light
Alireza Marandi’s Caltech team realized cellular automata using interacting light pulses. Reported in Light: Science & Applications, the demonstration explores a computing architecture built around local optical interactions, with potential applications in simulation and information processing.
Optical neural networks compress images before the camera detects them
Peter McMahon’s Cornell group demonstrated optical preprocessing for computer vision, retaining task-relevant information while greatly reducing the data sent to a camera. The Nature Photonics work points toward compact, energy-conscious imaging systems rather than simply moving all computation into electronics.
NTT plans expanded optics facilities for Ising-machine experiments
Yoshihisa Yamamoto’s PHI Lab announced plans to roughly double its optics-lab space, adding capacity for system-level coherent Ising-machine experiments and thin-film lithium-niobate devices. The expansion was intended to support both in-house research and university collaborations.
New Paper: Few-Cycle Vacuum Squeezing in Nanophotonics
The Nonlinear Photonics Laboratory at Caltech demonstrated an integrated nanophotonics platform based on lithium niobate to generate and measure squeezed states on the same optical chip.
Conference News
Coherent Network Computing Workshop 2022
The second Coherent Network Computing conference took place at Stanford, California from Oct 24-26, 2022.
Preprint
New pre-print: Programmable simulation of bosonic transport in optical lattices
The McMahon Lab, in collaboration with NTT Research, shows that photonics can make very large analog simulators that take advantage of the broad bandwidth of optics.
Publication
New Paper: Published in Nature Photonics Journal
A paper on femtojoule femtosecond all-optical switching in lithium niobate nanophotonics was published in Nature Photonics.
Preprint
New pre-print: Copositive Programming for Mixed-Binary Quadratic Optimization
The Stanford Autonomous Systems Lab released a preprint on solving mixed-binary quadratic programs using Ising solvers.